Cancer is a significant cause of illness-related deaths in the United States. It is estimated that approximately 17,000 malignant brain tumors are diagnosed in adults and 1,500 in children every year in the United States.
Brain tumors are usually lethal. Human brain tumors are typically classified as primary tumors and secondary tumors depending on their origin 1. Primary tumors originate in the brain and are classified according to the histological basis from which they are derived; for example, gliomas arise from glial tissue. Secondary tumors arise from metastatic primary cancers originating elsewhere in the body. The two chief sources of secondary brain tumors are lung cancer in the male and breast cancer in the female. The five year survival rate for primary brain tumors is only about thirty-five percent.
The most common therapy given to such victims is surgical resection. The normal-tumor boundaries for different primary and secondary brain tumors vary from fingerlike protrusions of tumor cells into normal tissues in glioblastoma multiforme to well-circumscribed nodules with possible surrounding edema in most secondary tumors. The most common initial therapy for primary and secondary brain tumors is surgical resection.
Many studies have shown that the degree of resection significantly influences the time to recurrence and the overall survival of brain tumor patients. Successful resection relies on complete removal of the tumor. Aggressive surgery is the brain is not acceptable, yet residual tumorous tissue left behind after resection is believed to be the main cause of morbidity.
Currently, surgical navigation systems and ultrasonography are used intraoperatively to help neurosurgeons locate brain tumor and maximize resection. Surgical navigation systems enable neurosurgeons to relate the position of a surgical instrument to structures present in preoperative computerized tomography (CT) or magnetic resonance (MR) images. However, CT or MR imaging may not delineate the exact brain tumor borders. Studies have shown that neoplastic cells can be found in brain tissue outside the apparent tumor margins defined by contrast-enhanced CT or MR imaging. More importantly, the accuracy of surgical navigation systems can be degraded by registration error and intraoperative brain deformation which may shift brain tumor borders in image space by more than a centimeter from their actual locations. Ultrasonography is able to detect brain tumors because of their hyperechoic characteristics. However, peritumoral edema is also hyperechoic, which hampers tumor and tumor margin identification. Thus, despite the applications of these technologies in neurosurgery, significant residual tumor mass is often found to be left behind in patients after craniotomy. Neurosurgeons also rely on visual inspection and/or on-site pathology to locate tumors and tumor margins. Visual inspection is subjective and often incorrect as the visual characteristics of many brain tumors mimic that of normal brain. In addition, on-site pathology is expensive and time-consuming. Hence, there is a need for an objective, intraoperative real-time system which is capable of accurately differentiating brain tumors from normal brain tissue, thus detecting tumor margins with sub-millimeter spatial resolution.
Optical spectroscopy, such as fluorescence spectroscopy, has been shown capable of detecting subtle changes in tissue architecture and biochemical composition associated with the progression of disease in near real-time. Optical spectroscopy has been successfully applied to detect disorders of various organ systems (e.g., cervix, skin, etc) both in vitro and in vivo. Several commercial systems are currently available for clinical diagnosis in the bronchus, cervix, etc. However, relatively few studies have addressed the diagnostic potential of optical spectroscopy in brain tumors. It has been reported that fluorescence peaks at 470, 520, and 630 nm emission were measured from human brain tissues in vitro at 360, 440, and 490 nm excitation, respectively. Others have observed significant differences in autofluorescence properties between normal and tumorous human brain tissues at 360 nm excitation. The results of these studies were inconclusive in terms of the effectiveness of autofluorescence spectroscopy alone for brain tumor demarcation.
Several investigators have used fluorescence dyes, such as 5-aminolevulinic (ALA), to enhance brain tumor detection. Low sensitivity of this method at the borders of infiltrating tumors has been reported as the fluorescence dye is not taken up by tumor cells where the blood brain barrier is intact. Moreover, ALA-induced fluorescence spectroscopy encounters additional problems including bleaching of fluorescence due to excessive or prolonged illumination. Consequently, dye-enhanced fluorescence spectroscopy may not be the ideal approach for brain tumor demarcation.
Diffuse reflectance spectroscopy is a fast, noninvasive method used to determine optical properties of a sample. It is typically obtained by illuminating a sample (e.g., tissue) with a broadband white light source. Because of the changes in structure and morphology at the cellular and sub-cellular level, certain optical properties of human normal brain tissues are different in certain respects from that of human brain tumorous tissues.
In one aspect, the invention is system for brain tumor margin detection comprising: a source of white light; a source of laser light at a wavelength of about 330-360 nm; a fiber optic probe coupled with the source of white light and the source of laser light so as to deliver the white light and the laser light to a working end of the probe; a spectroscope coupled with the fiber optic probe so as to receive autofluorescent and diffuse reflectance light returned from tissue contacted by the working end of the probe and provide a frequency spectrum of the returned light; a system controller including a processor coupled with the spectroscope and programmed to analyze the frequency spectrum to distinguish between light returned to the spectroscope from tumorous and from non-tumorous tissues.